Copilot in Excel for Finance Professionals

This course is for finance professionals -- FP&A analysts, accountants, controllers, and finance managers -- who are comfortable with core Excel skills (formulas, basic PivotTables, and formatting) but new to Copilot. No prior AI or scripting experience is required.

• Understand what Copilot in Excel can and can't do, and confirm their organization's licensing before relying on it
• Write clear, specific prompts that name the table, the task, and the format they want an answer in
• Explore and summarize an unfamiliar general ledger or expense dataset conversationally, without building a filter or sort by hand
• Generate, explain, and debug formulas -- including common errors like #REF!, #DIV/0!, #VALUE!, and #N/A -- using plain language
• Detect trends and outliers, and build and reshape PivotTables, for budget-vs-actual and transaction-level data
• Draft variance narratives and stakeholder summaries, and reframe the same finding for a CFO, a department head, and a peer analyst
• Run budget-vs-actual variance analysis end to end, narrate what-if scenarios, and use Copilot as a second set of eyes during month-end close
• Build and refine charts by describing the comparison they want to show, and know when a chart helps versus when a table is clearer

Foundations of Copilot in Excel

What Copilot in Excel does and doesn't do - Licensing on Business Basic and Business Premium - Preparing your data as Excel Tables - Where Copilot's answers come from -- grounded vs. general answers

Hands-on exercise: Students activate Copilot, convert a raw general ledger export into a properly structured Excel Table, and compare a Table answer to a plain-range answer on the same data.

Asking Excel Questions in Plain Language

The anatomy of a strong prompt -- context, task, and format - Being specific about scope - Exploring and summarizing a dataset conversationally - Refining prompts through follow-up - Recognizing when a prompt needs more specificity

Hands-on exercise: Students explore a regional expense ledger, ask directional questions to find the largest variances, and practice scoping a deliberately vague prompt down to a useful one.

Copilot-Assisted Formula Writing

Generating formulas from a plain-language description - Explaining an inherited formula step by step - Diagnosing common formula errors - Knowing when to trust a generated formula

Hands-on exercise: Students build a rolling 3-month average formula, debug four intentionally broken formulas -- one of each common error type -- and explain a Variance % formula they didn't write.

From Raw Data to Decision-Ready Insight

Trend and outlier detection - Assisted PivotTables - Written summaries for stakeholders - Tailoring the same data to different audiences

Hands-on exercise: Students build a PivotTable summarizing transactions by region and product line, draft a variance narrative for a department head, and reframe one finding for a CFO, a department head, and a peer analyst.

Applying Copilot to Core Finance Workflows

Budget vs. actual variance analysis - Lightweight forecasting support and its limits - Scenario ("what-if") narratives - Month-end close support - Building a repeatable weekly or monthly routine

Hands-on exercise: Students walk a budget variance story end to end, model the impact of a freight cost increase on COGS, and draft a monthly variance routine adapted to a report they actually own.

Visual Storytelling with Copilot

Chart recommendations by prompt - Iterating on chart design conversationally - Choosing when a chart helps and when a table is clearer

Hands-on exercise: Students build and refine a quarterly revenue chart across product lines, then compare a chart view to a table view of the same data to decide which one they'd actually attach to an email update.

Using Copilot Responsibly

Data sensitivity and permissions - Known limitations - Building a personal verification habit - Prompting best practices recap

Hands-on exercise: Students audit and rewrite one of their own earlier prompts, then commit to a single personal verification habit to carry forward.

Students should be comfortable with core Excel skills -- formulas, basic PivotTables, and formatting. No prior AI or scripting experience is required. Students' organizations must be licensed for Microsoft 365 Business Basic or Business Premium with the Microsoft 365 Copilot add-on activated for Excel.

In finance, the cost of an unnoticed error is higher than in most departments -- a plausible-sounding wrong number can move further before anyone catches it. In this course, an instructor is right there to answer the judgment-call questions as they come up, so "should I trust this formula?" or "is this variance real or a data issue?" gets settled in the room instead of after the fact.

Students also retain up to one year of access to a recording of the class and six months of post-class instructor support. That matters most for a course like this one, because the real test isn't the four hours in class -- it's the first time you run Copilot against your own month-end variance report.